PROJECT SUMMARY. Atrial fibrillation (AF) is the most common cardiac arrhythmia and markedly increases the risk for stroke and systemic embolism. Oral anticoagulation has been the mainstay of therapies to reduce the risk of stroke and systemic embolism in AF, but studies have consistently shown that between 30-50% of patients for whom oral anticoagulation is indicated are not treated with anticoagulants. Left atrial appendage occlusion (LAAO) represents a new and increasingly important device-based approach for preventing stroke and thromboembolism in AF patients. Randomized trials have evaluated LAAO for the prevention of stroke and thromboembolism in atrial fibrillation, but the trial results were inconsistent, leaving important gaps in knowledge regarding the safety and effectiveness of this procedure. The goal of this proposal, ?Safety and Effectiveness of Left Atrial Appendage Occlusion in Atrial Fibrillation (SAFELY-AF)?, is to leverage data from a large national registry of LAAO procedures to enhance our understanding of the use and effectiveness of LAAO in real-world practice. The National Cardiovascular Data Registry?s LAAO Registry represents the largest prospective registry of LAAO procedures worldwide. Through a novel collaboration with Boston Scientific and the American College of Cardiology (ACC), we propose a series of studies that will enhance our understanding of the use and effectiveness of LAAO in real-world practice. First, we will use patient-level data from the pivotal randomized trials of LAAO devices and data from the NCDR LAAO Registry to examine differences in patient characteristics and outcomes following device implantation. We will further identify patient, physician, and hospital predictors of adverse outcomes following device implantation and use this information to develop a risk score to determine major adverse events. This information will be incorporated into the ACC?s shared decision making (SDM) tool, developed to promote shared decision-making for patients and physicians regarding LAAO procedures. Finally, we will conduct comparative-effectiveness research, using cutting-edge computational approaches to characterize the association between different approaches to periprocedural antithrombotic and anticoagulant therapies and patient outcomes following device implantation.